This article describes the results of a study designed to test whether reactive criminal thinking (RCT) does a better job of mediating the crime → drug relationship than it does mediating the drug → crime relationship after the direct effects of crime on drug use/dependency and of drug use/dependency on crime have been rendered nonsignificant by control variables. All 1,170 male members of the Pathways to Desistance study (Mulvey, 2012) served as participants in the current investigation. As predicted, the total (unmediated) effects of crime on substance use/dependence and of substance use/dependence on crime were nonsignificant when key demographic and third variables were controlled, although the indirect (RCT-mediated) effect of crime on drug use was significant. Proactive criminal thinking (PCT), by comparison, failed to mediate either relationship. The RCT continued to mediate the crime → drug relationship and the PCT continued to not mediate either relationship when more specific forms of offending (aggressive, income) and substance use/dependence (drug use, substance-use dependency symptoms) were analyzed. This offers preliminary support for the notion that even when the total crime–drug effect is nonsignificant the indirect path from crime to reactive criminal thinking to drugs can still be significant. Based on these results, it is concluded that mediation by proximal reactive criminal thinking is a mechanism by which distal measures of crime and drug use/dependence are connected.